povertyand social exclusion in ruralareas€¦ · by contrast, rural development policiesare often...

European Commission

country:

FRANCEauthor: Luc Behaghel

POVERTY AND SOCIAL EXCLUSION

IN RURAL AREAS

Final Report

Annex I

Country Studies

24

Poverty and social exclusion in rural areas - Final report Annex I - Country Studies

FRANCE

Contents

Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

1. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

2. Main characteristics of rurality . . . . . . . . . . . . . . . . . . 28

3. Main characteristics of rural poverty . . . . . . . . . . . . . 31

4. Rural poverty and policies . . . . . . . . . . . . . . . . . . . . . 36

5. Poverty and groups at risk: case studies on significant groups at risk . . . . . . . . 43

6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

7. References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

AnnexEXAMPLES OF GOOD PRACTICES . . . . . . . . . . . . . . . . . . . . 53

This study is supported for under the European Community action programme to encourage coopera-tion between Member States to combat social exclusion (2002-2006). This programme is managed bythe Directorate-General for Employment, social affairs and equal opportunities of the EuropeanCommission.Neither the European Commission nor any person acting on behalf of the Commission may be heldresponsible for the use that may be made of the information contained in this publication.

http://ec.europa.eu/employment_social/spsi

Contractor: Fondazione Giacomo Brodolini

Contact person: Marilena Sacchetta

www.fondazionebrodolini.it

Authors of the Final Study: Paola Bertolini, Marco Montanari, Vito Peragine.

National Corrispondents: Lilia Abadjieva (Bulgaria), Luc Behaghel (France), Paola Bertolini (Italy),

Nikolaos Bouzas (Greece), Ruta Braziene (Lithuania), Patrick Commins (Ireland), Philomena De Lima

(UK), Oana Gherghinescu (Romania), Elvira Gonzalez (Spain), Florindo Ramos (Portugal), Karen

Refsgaard (Norway), Mateja Sedmak, Blaz Lenarcic (Slovenia), Elzbieta Tarkowska (Poland), Achim

Vanselow, Claudia Weinkopf, Thorsten Kalina (Germany), Gabriella Vukovich (Hungary).

Scientific Committee: Philomena De Lima, Marcello Gorgoni, Sabrina Lucatelli, Enzo Mingione,

Karen Refsgaard, Annamaria Simonazzi, Francesca Utili

© European Communities, 2008

Reproduction is authorised provided the source is acknowledged.

Executive Summary

In France, poverty rates are higher in rural districts than in small cities, and slightly higher than in large cities:

poverty and social exclusion are persistent phenomena in rural France.

This report stresses the diversity of rural France (section 2). In particular, the agricultural orientation of the econ-

omy is no longer predominant: rural areas with a residential orientation dominate, while those with a manufactur-

ing orientation are strongly represented. This diversity notwithstanding, poverty and social exclusion display some

distinctive features in rural areas (section 3): they are less correlated to long-term unemployment than in urban

areas, less concentrated upon lone-mothers families, and often connected to the lack of decent housing and to bad

access to services and employment, due to remoteness.

Policies impacting poverty and exclusion in rural areas (section 4) take three important orientations: they are rarely

specifically targeted on rural areas; when they are, they disproportionately target the agricultural sector; moreover,

they increasingly tend to select territories on their ability to produce local development strategies.

Are these good orientations? The analysis leads to balanced answers:

➣ While the lack of focus on rural characteristics may miss problems that are specific to the rural poor – e.g.,

their difficulty to activate their rights –, existing policies appropriately try to strengthen the links between

rural and urban areas. This acknowledges the growing role played by these links, in particular through the

residential function or rural areas.

➣ The “sectoral resistance” of French policies in favour of the agricultural sector is at odds with the declining

role of agriculture in rural areas. However, forestry and agribusiness still have a significant weight in rural

employment, and poverty among farmers is a real and persistent problem.

➣ Last, the selection of territories is explicitly conditioned by their ability to group in order to produce docu-

ments about their local development strategies. This bottom-up approach arguably increases the level of ini-

tiative. However, existing studies also suggest that more dynamic territories that have the appropriate human

resources may disproportionately benefit from the funds, thus widening the gap between territories.

An important result of our review is that we lack true impact evaluations of the different policy measures. Clearly,

counting the number of recipients or reporting the total spending is not enough. Several micro measures – for

instance, subsidizing the investments of micro firms, or the diversification investment of farming households –

would deserve rigorous econometric evaluations, and could even be evaluated by controlled experiments using ran-

domized trials. Concerning more macro measures, we now have evaluations of objective 5b programs and of fis-

cal exemption measures granted to rural areas: the impact on economic activity and employment is not significant,

begging the question of whether these measures are sufficiently focused and efficiently administered.

The case study of poor farmers (section 5) illustrates the persistence of poverty within this group. Poverty among

farmers has highly specific features in the French context. Whereas poverty and exclusion is generally related to

unemployment or underemployment, a substantial share of farmers (22%, according to the most recent numbers)

constitute a group of full-time working poor. This specificity can be explained by a conjunction of factors: the

structural decline in the price of agricultural goods; the specific adjustment process of the agricultural sector

(adjustments take place through prices – revenues – rather than through quantities – employment –; correlatively,

in most cases, farming remains a lifetime job); the relative independence with regard to national social policies (as

exemplified by the low access to the minimum income, the RMI).

This begs the question of the efficiency of public policies concerning poor farmers. We find that the first pillar of

the CAP probably helps reducing poverty occurrence among French farmers by providing a safe, though limited,

income to all. The impact of rural development policies (including those of the CAP second pillar) on poverty

among farmers remains uncertain, but is probably limited.

25


FRANCE

Given the diversity of the needs, the practices of specialized institutions and local associations may be appropriate

responses to specific situations of poverty. We analyze two such practices (section 6): the measures put into place

by the Mutualité Sociale Agricole to detect and prevent risks of social exclusion early; and the role played by two

associations in Lorraine as intermediaries on the labour market for unskilled workers. We show that the relative

success of these practices is based on the adaptation of national measures to specific territories and populations.

To summarize and tentatively conclude, this report shows that, after having been neglected for decades, the terri-

torial dimension of poverty and social exclusion has been widely recognized in France, as well as the need for

locally specified policy measures. The risk now seems that these measures may be insufficiently focused, unequal-

ly activated by the least organized territories and groups, and not everywhere efficiently implemented. More effi-

cient policy interventions could be based upon less ex ante monitoring, better ex post evaluation – including true

impact evaluations of some key policy measures – and a real involvement of the State in providing tools and

expertise, and in disseminating to the whole territory the lessons learnt by local, experimental projects.

26


FRANCE

1. Overview

The strong heterogeneity of rural areas and the various types of rural poverty raise the question of the public poli-

cies that need to be developed. A recent policy evaluation report (Perrin, 2003) stresses the limit of policies target-

ed toward narrowly defined sectors (in particular, agriculture) and the need to involve local players more in order

to adapt local policies to local needs. But in a different perspective (mostly based upon evidence of stigmatisation

and resources spreading in urban areas), Maurin (2005) questions the whole logic of policies targeting territories:

should they not instead target specific groups within territories?

If targeting specific groups or individuals is the road to be followed, then the findings of Pagès (2004) need to be

borne in mind: exclusion in rural areas is, more often than in urban areas, linked to the difficulties or reluctance of

poor people to apply for the benefits to which they are entitled, and the difficulties of social workers to find them.

The orientations currently taken by the policies that may affect poverty and social exclusion in rural areas are char-

acterized by two features:

• A still strong sectoral bias toward agriculture;

• The predominance of the objective of developing employment through increased economic initiatives (con-

sistently with the Lisbon strategy).

The net impact of these measures on employment in rural areas is not well-known – and this is probably the first

finding of our analysis: the lack of true impact evaluations. One of the rare macroeconometric analyses available

suggests that the impact could be quite mixed (Lofredi and Schmitt, 2006). This justifies the stress given by the

European Commission not only in favour of ex ante but also of ex post evaluations of the existing programs. Taking

net job creation as the measure of impact makes certainly sense; however, it would also progressively be necessary

to translate this into an impact on the income distribution.

27


FRANCE

2. Main characteristics of rurality in France

2.1. Definitions of rural areas

Administrative documents currently use three alternative and complementary definitions of rural areas:

• The OECD classification of predominantly rural / urban / intermediary areas is used in the most recent doc-

uments (e.g. diagnostic in the National Strategic Plan for Rural Development (PDRH, 2007)

• The Zonage en aire urbaine (ZAU) definition is perhaps the most frequently used . It has been established

by INSEE and extended to rural areas [Insee and Inra, 1998]. The ZAU defines rural areas and peri-urban

areas by their degree of connection with urban centers, in terms of employment.

The classification is done at the municipality (commune) level which corresponds to LAU 2. Specifically:

- urban centers (communes du pôle urbain) are defined by the concentration of more than 5,000 jobs;

- peri-urban areas (communes périurbaines) have less than 5,000 jobs but have more than 40% of their res-

idents commuting daily to urban centers or to other municipalities of the peri-urban area;

- rural areas (communes rurales) cover the remaining of the French territory.

Urban center(s) and their peri-urban areas constitute urban areas (zones urbaines). Rural areas themselves

are subdivided into three sub-categories:

- rural areas under weak urban influence that have more than 20% and less than 40% of their residents com-

muting to an urban area;

- rural centers (pôles ruraux) with 2,000 to 5,000 jobs, and their fringes;

- remote rural areas (rural isolé).

• Beside the ZAU, the old definition of rural municipalities (dating back to the 1846 census) is still in use:

a municipality is said to be rural if it has fewer than 2,000 inhabitants leaving in the same agglomeration

(i.e. with less than 200 meters between two houses).

Policy zoning schemesThe National Report on Strategies for Social Protection and Social Inclusion (PNAI, 2006) rarely mentions rural

areas specifically. The difficulties of urban fringes are more frequently mentioned than those of remote rural areas.

By contrast, rural development policies are often specified according to different zoning schemes. A report on

rural development policies (Perrin, 2003) has insisted upon the fact that these zoning schemes overlap only par-

tially. For instance, areas eligible for objective 5b and objective 2 programs differ generally from those eligible for

fiscal exemption (Zones de revitalisation rurale, ZRR), although population density was one of the criteria used in

both cases. The partial overlap of zoning schemes is sometimes viewed as a sign of inconsistency; but it can also

be a way to smoothen the boundary effects that each individual scheme creates.

Rural zoning schemes have come under sharp criticism over the past ten years: they were viewed as too complex,

arbitrary and useless given the existence of alternative targeting schemes (Mondot et al., 2006). Instead of target-

ing broad regions or broad categories of space, rural development policies have tended to target individual recipi-

ents (e.g., farmers who sign a contract with the administration – Contrat territorial d’exploitation replaced in 2002

by the Contrat d’agriculture durable) or to fund local projects (e.g. LEADER program and pôles d’excellenceruraux). However, zoning schemes remain broadly used: for instance, the Zones de revitalisation rurale have been

recently redesigned without major changes.

Comparison between classificationsThe OECD classification aggregates communities defined as rural or urban on the basis of their population densi-

ty into predominantly rural and predominantly urban regions.

By comparison, the French Zonage en aires urbaines does not barely rely upon population density, but also upon

job concentration and home-to-work commuting flows. This has the advantage of focusing upon the connections

28


FRANCE

between urban and rural areas, whose importance has greatly increased over the past decades. Nonetheless, the

OECD classification is more easily available and allows for international comparisons, whereas the Zonage enaires urbaines is derived from a complex algorithm.

The “rural index” developed by this study for cross-country comparisons cumulates information on the population

density (excluding the largest urban area) and on the share of gross value added in agriculture. It thus approaches

criteria that have been used at yet another administrative level (group of municipalities called cantons) in the selec-

tion for objective 5b and objective 2 programs.

When using these various classifications, it is necessary to keep in mind the heterogeneity of rural areas in France.

This is obvious from a physical geography viewpoint, with a sharp contrast between mountainous, sometimes quite

remote areas, and the plains. A specific zoning is devoted to mountains (zone de montagne).

2.2. The different typologies of rurality in France

Besides geographical classifications, other space categories based on the economic orientation, the employment

profile and the access to public facilities are probably more relevant to the study of poverty. Such categories have

been recently proposed (DATAR, 2003): rural France has been divided into 1745 areas, organized around cities

with fewer than 30,000 inhabitants. Each “living area” (bassin de vie) is defined as the space in which most inhab-

itants look for employment and for key public facilities (for health, education, shopping, etc.). Different typologies

can then be proposed, to characterize the degree of autonomy of the areas (in terms of access to job and to public

facilities) and their economic orientation. In particular, three broad categories of economic orientations have been

defined: residential (48.7% of French territory and 21.9% of the population), agricultural (8.1% and 2.6%) and

manufacturing (17.8% and 10.1%). This typology shows the limited role played by the agricultural orientation, and

the persistent role manufacturing in French rural areas.

This classification has been used in a slightly modified version by the latest National strategic plan for rural devel-

opment (PDRH, 2007; figure 1).

Figure 1: Classification of French municipalities according to their economic orientation

2.3. Main social and economic problems in rural areas in France

Although this is partly subjective, it is useful to list the main social and economic problems of French rural areas.

We base this list on the diagnostic made by the French National strategic plan for rural development (PDRH,

2007):

29


FRANCE

• Unemployment

The high level of unemployment is common to rural as well as urban areas (while intermediary areas fare

better).

One specificity of rural areas is the low employment rate of women (36.4% vs. 41.2% in intermediary

areas). A higher fraction of women is looking for a job in rural areas (14.4% vs. 12.6%) or not participating

to the labour market. This can in particular be due to the difficulties with child care (lack of child-minding

services, transportation problems).

• Spatial segregation and service availability

Since 1975 the demographic decline of French rural areas has stopped and the population of rural areas has

started growing again on average. However, cities also continue increasing their area of influence, and this

creates increasing pressure on the price of land. Part of the population of households with a lower income

has tended to move further away from the cities; as a result, there is a clear decreasing gradient in income

as one moves away from cities. This is aggravated by the lower availability of services (shops, child-mind-

ing services, public transportation).

• Population ageing

Population ageing is accentuated in rural areas. In 1999, more than 25% of the rural population was older

than 60. For France as a whole, the proportion was only 21% (PDRH, 2007).

• Heterogeneity

Although underscoring these general trends is useful, it is probably even more useful to underscore the

diverging socio-economic trends of different rural areas in France. Following the diagnostic made by the

CAESAR research team, and as illustrated by figure 2, one can distinguish three different trends in French

rural areas:

- Rural areas that benefit from the proximity of dynamic urban areas and that have a strong economic poten-

tial (Rhônes valley, South-East, Atlantic coast, Ile-de-France).

- Rural areas “in search of new equilibrium” (North, East).

- Fragile rural areas, characterized by a demographic and economic decline (Auvergne, Limousin).

Figure 2: Performance of rural areas

30


FRANCE

3. Main characters of rural poverty in France

3.1. Measurement of poverty in France

The national definition of the poverty rate has been aligned on the Eurostat definition: poor households have equiv-

alised (per unit of consumption, OECD scale) disposable income below 60% of the national median. Regional

thresholds are not normally used.

This statistic is computed from the SILC data as well as from a variety of sources: administrative fiscal data

(Revenus fiscaux), consumption surveys (Budget des familles), and other household surveys.

However, recent research (see in particular the special issue of Economie et statistiques introduced by Verger,

2005) as well official reports by the national observatory of poverty and exclusion (Observatoire national de lapauvreté et de l’exclusion sociale, ONPES) have stressed the necessity to use a broader range of indicators to

account for the multiple dimensions of poverty. The ONPES has thus proposed a list of ten indicators of poverty

and exclusion:

• five poverty indicators: the poverty rate (as defined above), two indicators of poverty persistence and two

indicators of poor living conditions;

• four indicators of exclusion, relating to:

- employment (rate of unemployed workers who have exhausted their unemployment benefits);

- health (rate of non access to health services by lack of resources);

- education (rate of young people leaving school with low achievement);

- housing (rate of unsatisfied housing demands);

• one indicator of inequality (the interdecile ratio).

Poverty rates in urban and rural areasThe poverty rate is available separately for rural and urban areas.

According to ONPES (2006), poverty tends to become more and more an urban phenomenon. Between 1996 and

2002, the poverty rate has grown in large cities (above 200,000 inhabitants) and decreased in small cities and rural

areas. However, in the cross-section, rural districts (below 2,000 inhabitants) still have the highest proportion of poor

people (slightly above 25%, compared to about 24% in cities above 200,000 and to about 13% in cities below 20,000).

It should be noted, however, that such comparisons of poverty rates are not robust to changes in the methodology:

in particular, poverty rates decrease significantly in rural areas if they are corrected for the fact that many rural

households own their home and do not pay rents. Moreover, these poverty rates do not account for local variations

in the price of goods and services: in that sense, they may not capture differences in real income well.

3.2. Rural poverty in France

The analysis of rural poverty and exclusion involves a variety of stakeholders – local and national administrations,

associations, local politicians, social workers, and researchers. In particular, the mid-term evaluation of the last

National Rural Development Plan (2000-2006) has involved stakeholders at the regional and national levels in a

long-ranging process (Lacombe et al., 2006).

The “rediscovery” of the rural poor in official documentsThe past few years have witnessed several policy reports on spatial inequalities, poverty and segregation. Until

recently, most reports have focused upon urban areas, where problems are often perceived as the most acute.

However, poverty in rural areas has been put forward by recent reports, in particular by the Observatoire nationalde la pauvreté et de l’exclusion sociale (ONPES, 20061). This “rediscovery” of rural poverty in policy reports

reflects the increased attention to local poverty situations, partly induced by the decentralization of social policies.2

31


FRANCE

This decentralization is partly based on the premise of a strong variability across regions, but this diversity has not

been well documented.3 Recent reports have therefore tried to use existing administrative data to document region-

al variations and to synthesize the existing local studies (see Mansuy and Pallez, 2002). The renewed attention to

rural poverty is partly a result of this effort: as argued by Péchoux (2006), it probably reflects better statistical cov-

erage more than changes in the phenomenon itself.

The last National Report on Strategies for Social Protection and Social Inclusion (NPAI, 2006) mentions two spe-

cific groups at risk:

• the seasonal agricultural workers;

• the young unemployed.

Main at-risk groups identified in the scientific literatureDavezies and Korsu (2002) show that older people (over 65) are overrepresented among the rural poor and lone

parents and single individuals are underrepresented compared to what happens in urban areas. Moreover, rural

poverty is less correlated to long-term unemployment than urban poverty.

In an analysis of the evolution of spatial income inequality during the past two decades, Behaghel (2006) shows

how the contrasts between the rural and urban populations (share of farmers and blue collars, but also family struc-

ture, labour force participation and employment) can statistically explain a significant share of the differences in

income per consumption units; the faster rise of unemployment in urban areas is one of the major forces driving

the closing of the rural-urban income gap between 1984 and 2002.

Beyond these general characterizations of rural poverty, several studies have focused upon specific groups

(younger or elderly people, blue collars, farmers, etc.). In particular, several studies have consistently shown the

persistence of poverty among farmers, despite the steady decline in the number of farms that could have implied

the survival of the most profitable ones (Jégouzo, Brangeon and Roze, 1998; Blanc and Perrier-Cornet, 2001;

Guillaume, 1999). Blanc and Perrier-Cornet (2001) show that even when averaged over three years (1996-1998),

farm income remains negative or below the minimum wage in about one farm out of five. A vast majority of those

farmers complete their income with other activities, but the additional income is low in most cases. In addition, the

opportunity for additional revenues appears to decrease with the distance to labour markets: the probability of rely-

ing on social assistance (Revenu minimum d’insertion, RMI) increases with the distance to large cities.

Davezies and Korsu (2002) suggest four factors to explain the different levels of poverty in France. The first one is

the level of economic development: high levels of economic development help explain the low poverty rates in

regions like Alsace, Ile-de-France and Rhône-Alpes. Second, the “form” of development comes into consideration: in

particular they argue that dense networks of small or medium-size businesses resist better to economic shocks than

regions with large businesses and few entrepreneurs. Third, the economic orientation matters: mono-industry manu-

facturing regions (Lorraine, Pas-de-Calais) have suffered from the decline of manufacturing in France. Last, the repar-

tition of poor people is largely driven by migrations: some regions tend to “export” poor people (e.g. Lorraine) while

others attract them (e.g. Languedoc). On average, rural areas do not fare well on these four factors: “Generally speak-

ing, rural areas – even more so in the Western part of the country – cumulate delayed economic development, weak

demographic and economic dynamics and poverty” (Davezies and Korsu, 2002, p. 236). However, the authors recog-

nize that this is not sufficient for causal analysis: what are the mechanisms behind these four factors?

Recent work in economic geography has attempted to deepen the analysis while dealing specifically with rural

areas. In particular, the low density of the population in rural areas has been put forward as a specificity of rural

areas that helps explain their economic specialization and some features of unemployment. Blanc, Aubert and

Détang-Dessendre (1999) stress the impact of higher matching frictions on rural labour markets (difficulties for

firms to find the desired worker, and symmetrically for worker). These frictions help explain why firms looking

for a skilled workforce rarely settle in rural areas while those looking for a stable, unskilled workforce might be

attracted by the lower equilibrium wages. This might have consequences on the share of “working poor”: the

descriptive results of Davezies and Korsu (2002, p. 215) show that the share of low wage workers is somewhat

higher in rural areas. Ongoing work by Détang-Dessendre and Gaigné (2006) analyzes the consequences of match-

ing frictions on unemployment duration. It tends to show that, ceteris paribus, unemployed workers in rural areas

have more difficulties finding a job when the density of the labour market gets lower.

The mechanisms of migrations and segregation have been studied from a variety of perspectives – geographic, eco-

nomic, and sociological – but with a strong focus on cities and their close urban fringes. However, the same kind

of mechanisms seems to apply to distant peri-urban areas and to rural areas. Guilluy and Noyé (2004) contend that

the France of poor people is the France of peripheries – distant peri-urban areas and rural areas – to which house-

holds migrate, either to access home ownership, for those with sufficient resources, or in search of the lower cost

32


FRANCE

of living. In both cases, migrants are often caught in the trap of scarce public facilities and insufficient transporta-

tion systems that disconnect them from job opportunities. The specific case of poor urban people migrating to

remote rural areas to take advantage of lower cost of living has been stressed by Maclouf (1986) who called them

“refugees from cities” (réfugiés de la ville).

3.3. Multi-dimensional analysis of poverty and social exclusion in France

Geographical location Poverty prevalence varies greatly across areas, even within a specific type of space (rural or urban). However,

poverty rates are not computed locally, but an approximation is given by the share of households with low fiscal

income (first decile). Poor households are concentrated in the South-West part of France (West of a line going from

Saint Brieuc to Dignes) and in the North (Davezies and Korsu, 2002).

As stated above, poverty tends to become more and more an urban phenomenon (ONPES, 2006). Between 1996

and 2002, the poverty rate has grown in large cities (above 200,000 inhabitants) and decreased in small cities and

rural areas. However, in the cross-section, rural districts (below 2,000 inhabitants) still have the highest proportion

of poor people (slightly above 25%, compared to about 24% in cities above 200,000 and to about 13% in cities

below 20,000)4. Such comparisons of poverty rates are not robust to changes in the methodology: in particular,

poverty rates decrease significantly in rural areas if they are corrected for the fact that many rural households own

their home and do not pay rents.

Physical infrastructure, accessibility (roads, railways, etc..), housing quality and utilitiesIn January 2004, 98.4% of the population lived at less than 45 minutes from one at least of the three following trans-

portation infrastructures: a railway station with more than 50 departures per day, a highway or an airport. The rest

of the population lives in mountainous, low density areas. As measured by the ORATE program, accessibility is

lower in mostly rural départements of the Massif Central or in the West of France (DATAR, 2005, pages 95 and 96).

Poor households own their house more frequently in the South-West part of France (West from a line going from

Cherbourg to Montpellier) and in rural areas: 50% of households in the lowest income decile own their house in

rural areas, compared to 18% in the Paris region (Davezies and Korsu, 2002). However, housing quality is often

low: 40% of non decent housing is found in rural areas (PNAI, 2006).

The distance to services such as mail offices, police station (gendarmerie) in unequipped municipalities has

remained roughly stable in France in past decades. It was 5.6 km and 7 km, respectively, in 1998.

Access to health care and long term careThe supply of health care can vary significantly from one region to the other. For instance, the number of doctors per

inhabitants is 70% higher in Ile-de-France and Provence-Alpes-Côte d’Azur than in Picardie. In 2004, the ObservatoireNational de la démographie des professions de santé established a typology showing that 1.6% of the population suf-

fered from a low access to health care, mainly in the North of France and in overseas regions (DATAR, 2005).

Environmental problemsEnvironmental problems take very different forms in rural and urban areas: air pollution in cities, water pollution

in some agricultural regions such as Bretagne (DATAR, 2005).

Types of activities and labour market structureHowever, other space categories based on the economic orientation, the employment profile and the access to public

facilities are probably more relevant to the study of poverty. Such categories have been recently proposed (DATAR,

2003): rural France has been divided into 1745 areas, organized around cities with fewer than 30,000 inhabitants. Each

“living area” (bassin de vie) is defined as the space in which most inhabitants look for employment and for key public

facilities (for health, education, shopping, etc.). Different typologies can then be proposed, to characterize the degree

of autonomy of the areas (in terms of access to job and to public facilities) and their economic orientation. In particu-

lar, three broad categories of economic orientations have been defined: residential (48.7% of French territory and 21.9%

of the population), agricultural (8.1% and 2.6%) and manufacturing (17.8% and 10.1%). This typology shows the lim-

ited role played by the agricultural orientation, and the persistent role manufacturing in French rural areas.

33


FRANCE

Farming, farmers and agricultural employees subsistence economy, unpaid family workersAs stated above, several studies have consistently shown the persistence of poverty among farmers, despite the

steady decline in the number of farms that could have implied the survival of the most profitable ones (Jégouzo,

Brangeon and Roze, 1998; Blanc and Perrier-Cornet, 2001; Guillaume, 1999). Blanc and Perrier-Cornet (2001)

show that even when averaged over three years (1996-1998), farm income remains negative or below the mini-

mum wage in about one farm out of five. A vast majority of those farmers complete their income with other activ-

ities, but the additional income is low in most cases. In addition, the opportunity for additional revenues appears

to decrease with the distance to labour markets: the probability of relying on social assistance (Revenu minimumd’insertion, RMI) increases with the distance to large cities.

More recently, the attention has been drawn upon the difficulties faced by seasonal agricultural workers (NPAI,

2006).

Family compositionLone parents and lone men and women are less frequent in rural areas (table 1).

Table 1: Family composition in urban and rural areas

Type of family Urban areas Rural areas

Lone man 16.0 10.5

Lone woman 22.6 16.2

Couple without children 21.4 11.8

Couple with one child 11.1 11.8

Couple with two children 11.1 11.8

Couple with three children or more 4.5 3.5

Lone parent family 4.4 1.6

Other households 8.9 17.0

Total 100 100

Source: Fall and Verger (2005), based on the ECHP (2001)

Even when controlling for this structural difference, the risk of poverty is somewhat lower among rural lone par-

ents / lone men or women than among their urban counterparts (Davezies and Korsu, 2002).

Age structureThe age pyramid is skewed toward older people in rural areas (table 2).

Table 2: Age structure of households

Age of the household head Urban areas Rural areas

<30 7.0 6.1

30-39 18.0 20.3

40-49 21.5 15.4

50-59 15.6 15.0

60-69 12.1 14.8

>70 25.8 28.4

Total 100 100

Source: Fall and Verger (2005), based on the ECHP (2001)

34


FRANCE

Moreover, the risk of poverty is above average the corresponding national average in rural areas among households

whose head is aged below 25 or above 60, whereas it is close to the national average for those aged 25 to 59.

Taken together, these two facts explain why poverty in rural areas predominantly concerns households whose head

is above 60: they make up to 50% of the poor, compared to 34% on average in France (Davezies and Korsu, 2002).

However, this specific feature of rural poverty tends to decline, in particular due to improving retirement conditions.

Access to educationThe average distance to primary schools has increased from 1.6 km in 1980 to 3.7 km in 1998 for (rural) munici-

palities which do not have a primary school, while the corresponding average distance to secondary schools has

decreased over the same period, from 8.3 km to 7.6 km. As of 1994, 54% of municipalities in remote rural areas

(as defined by the Zonage en aires urbaines) had a primary school. The decline in the number of rural primary

schools is linked to a strategy of grouping schools over two to six municipalities when the number of pupils

becomes too small.

GenderLone women and lone parent families are less frequent in rural areas (see table 1 above). Overall in France, pover-

ty rates are higher among women than men: respectively 12.6% and 11.8% in 2002. This gender gap increases with

age: the poverty rates are 11.6% and 9.7% above age 64 (ONPES 2006). This difference, however, is not docu-

mented as specific to rural areas.

EthnicityFrench statistics do not produce data on ethnicity as such.

MigrationRecent decades have witnessed a reversal of migration flows from urban to rural areas. However, the population

of urban areas keeps growing faster due to higher natural growth (fertility is higher). There are strong disparities

from one region to the other (Perrin, 2003, page 31).

Foreign immigration mostly concentrates in urban areas.

As noted by Davezies and Korsu (2002), the repartition of poor people is largely driven by migrations: some

regions tend to “export” poor people (e.g. Lorraine) while others attract them (e.g. Languedoc). The extent to

which migration of poor urban people to the rural areas is a significant source of rural poverty is still discussed.

As noted above, Guilluy and Noyé (2004) contend that the France of poor people is the France of peripheries –

distant peri-urban areas and rural areas – to which households migrate, either to access home ownership, for those

with sufficient resources, or in search of the lower cost of living.

CommutingCommuting has been extending quickly over the past decades. The share of the population living in municipalities

with more than 40% of residents commuting to the urban center has risen from 9% in 1982 to 18% in 1999

(Behaghel, 2006).

CrimeRural areas usually witness lower crime rates than rural areas (predominantly rural regions like Limousin, Auvergneand Bretagne had a crime rate below 45 per 1000 inhabitants in 2004, compared to a national average of 64).

Notes

1 See chapter 3 about “Territories and poverty”.2 E.g., the distribution of minimum income – RMI – by local authorities – the départements – since 2003.3 As noted in ONPES (2006), poverty rates are not available at the département level.4 These figures are based on relative poverty measures.

35


FRANCE

4. Rural poverty and policies

4.1 Rural poverty and social policies in France

The last National Report on Strategies for Social Protection and Social Inclusion (NPAI, 2006) sets three priorities

for the 2006-2008 period:

1. Increasing the access to employment for those who are the furthest away from employment;

2. Improving the integration of young people into the job market;

3. Developing the supply of social housing.

These priorities are transversal and are not specially targeted at rural areas. Some specific implications for rural

areas are nonetheless mentioned in the report:

1. Concerning access to employment, the report stresses the need to encourage alternatives to public trans-

portation for the unemployed in remote rural areas;

2. The case of young rural people and their integration into the job market is not addressed specifically;

3. Concerning social housing, the report stresses the lack of decent housing for seasonal agricultural workers

as well as the bad housing conditions of a significant fraction of rural residents.

4.2 Rural poverty and main policies affecting rural areas

Assessing the effectiveness of European / national policies to reduce poverty and social exclusion in rural areas

requires defining criteria of achievement and agreeing on evaluation methods. Evaluations should ultimately

include an evaluation of the policies’ impact on the outcome of interest (see for instance the guidelines given by

the European Commission that recalls the distinction between the output, the outcome and the long-run impact of

a policy – European Commission, 1999).

The econometric and statistical literature on program evaluation underscores the difficulty of true impact evalua-

tions outside experimental contexts and sets the standards for analyzing the causal impact of policies (Heckman,

Lalonde and Smith, 1999). The goal of impact evaluation is precisely stated in the framework of the Rubin model

(Rubin, 1974). The causal effect of a program, for a given individual (or entity), is defined as the difference

between the outcome for this individual in the presence of the program and the outcome for the same individual in

the absence of the program (the counterfactual). The same individual is never observed at the same time in the

presence and in the absence of the program; in other terms, the counterfactual is never directly observed, and it has

to reconstituted by the evaluator. This requires making assumptions that are more or less credible. The goal of

impact evaluation is to make inference that relies upon minimal, plausible identifying assumptions. The “gold stan-

dard” for impact evaluation is thus given by controlled social experiments which are based on randomized trials:

recipients and non-recipients are randomly chosen so that the control group – the non-recipients – are on average

statistically identical to the treatment group – the recipients – and therefore constitute a valid counterfactual under

minimum assumptions.

By contrast, the indicators often used are rather output indicators that do not allow for impact evaluation. Take the

measure 311 in the National Strategic Plan for Agricultural and Rural Development (PDRH, 2007). This measures

aims at facilitating the diversification of farming households towards non agricultural activities by subsidizing the

investments made by the households. The indicators monitored are the number of recipients and the total volume

of investment subsidized. This measure of output is not sufficient to assess the causal impact of the measure on the

share of farming households that diversify their activity. Indeed, the measure may be used by those who would

have developed the activity anyway, crowding out those who need the credits more, and having no aggregate

impact on diversification. Typically, on such micro measures, better impact evaluations are feasible, e.g. random-

izing the timing of the introduction of the subsidies, or with variations in the rate of public funding. This would

allow to infer what the measure actually changes to multiactivity of farming households, hence to poverty.

A first result of our review and assessment of public policies related to poverty and social exclusion in rural areas

36


FRANCE

is the lack of evaluations that meet the standards of true impact evaluations. And a first recommendation is to antic-

ipate the need of impact evaluations, not only, as is already well done by EC programs, by requiring ex ante, expost and in itinerae evaluations, but also, whenever possible, by designing the policy and its implementation in

ways that allow for robust impact evaluations.

A second difficulty is the multidimensional nature of poverty and social exclusion. Individual policy measures do

not try to reduce all aspects of poverty. They are combined into policies that entail a variety of measures. Moreover,

policies that matter for poverty and social exclusions in rural areas are not always intended to fight poverty; when

they are, it is not always their only objectives. It is thus necessary to disentangle the impact on poverty and social

exclusion in rural areas of a variety of measures, included in a variety of policies that also have different objec-

tives.

This being said, our assessment of existing measures will follow two main approaches:

1. The first one is to look for measures and existing policies for which we can come the closest to impact eval-

uations in terms of variables that are important drivers of poverty and exclusion (e.g. employment).

However, focusing only on these measures would run the risk of assessing a specific group of measures at

the expense of other measures that are harder to evaluate, but that may be shaping poverty and exclusion as

well. Therefore, this first approach provides useful “zooms” through robust ex post assessment of specific

measures, but is not sufficient to get a comprehensive picture of the reality.

2. The second approach is an a priori assessment of policy designs. Are they targeted towards groups that are

most at risk of poverty? Are the various measures combined into a consistent policy? Such ex ante evalua-

tion strongly relies upon the evaluators’ knowledge (or beliefs) about what works and what does not work.

Furthermore, it does not tell whether the policies were implemented according to the initial design. It thus

needs to be complemented by the first approach.

By convenience, our assessment will be divided into two groups of policies. Section 1 will assess agricultural and

rural development policies (in particular, the Common Agricultural Policy, CAP). Section 2 will address structur-

al policies.

4.2.1. Agricultural and rural development policies

The French Strategic Plan for Agricultural and Rural Development provides an up-to-date reference to agricultur-

al and rural development policies for the 2007-2013 period. We focus on the measures for metropolitan France

detailed in the PDRH (2007).

We first review the overall policy structure, and then focus upon two specific measures: the past experience of

objective 5b programs (an example of an integrated rural development program) and the fiscal exemption meas-

ures concerning the Zones de revitalisation rurales (ZRR).

➣ Overall policy structure: recent trends and the 2007-2013 National plan

Sectoral resistance: the focus on agricultureThe allocation of financial resources shows what Guérin (2006) calls the “sectoral resistance” of French policies

in rural areas: they remain predominantly targeted at the agricultural and agribusiness sectors.

The appendix details the overall public spending for rural areas and agriculture, adding national and EU funds. As

of 2005, a few facts stand out. The interventions on agricultural markets (the domain of the 1st pillar) represent

more than 10 billion euros out of a total of 12.6 billion. Moreover, that share tends to grow: between 1990 and

2005, it rose from 78.8% to 82.3%.

Figure 3 details the evolution of the funds allocated to rural development policies. The total spending has actually

declined sharply from 1994 to a low point in 2001, before rising and stabilizing. The structure of spending has also

evolved: agri-environment measures and transfers to less favoured areas have been progressively substituted to

subsidies devoted to young farmers starting up in farming.

37


FRANCE

Figure 3: Evolution of public funds allocated to rural development in France

Source: Dervieux (2007)

The recent French Strategic Plan for Agricultural and Rural Development confirms these trends. Within the 2nd pil-

lar, France has chosen to devote the minimum (10%) of EU credits to rural development policies. And within these

rural development policies, a significant share goes to the agricultural sector (in particular, agri-environment meas-

ures and measures in favour of less favoured areas).

The question of whether this sectoral bias in favour of agriculture is efficient and sustainable is debated. As the

Strategic Plan recalls, agriculture, forestry and agribusiness still have a significant weight in rural employment

(1.2m jobs in agriculture and 0.6m in agribusiness). Agribusiness represents 10% of the manufacturing sector.

Nonetheless, as discussed by Perrier-Cornet (2006), in the most likely scenarios, the economic role of agriculture

will continue declining, putting the current bias of public policies into question.

The 2007-2013 rural development planFollowing the EU guidelines, the PDRH (2007) is divided into 34 measures classified along four axes:

• Axis 1: Improving the competitiveness of the agricultural and forestry sectors: 31% of EU funds;

• Axis 2: Improving the environment and countryside: 58% of EU funds;

• Axis 3: Improving the quality of life in rural environment and encouraging diversification: 10% of EU

funds;

• Axis 4 (Leader): Building local capacity for employment and diversification (5% of EU funds spread over

the first three axes).

We focus our review on axis 3 to complement the review of the sectoral measures reviewed in the case study about

poor farmers, and provide a brief ex ante assessment of some key measures:

- Measure 311 provides subsidies to facilitate the diversification of farming households towards non agricul-

tural activities. This is consistent with the finding that multiactivity is a way of exiting poverty for farming

households. The focus of the measure (agritourism, tourist accommodation, services, direct sale of farm

products) is also consistent with the evidence that the residential function is an important source of growth

for rural areas: for instance, Davezies (2001) claims that revenues brought by new residents (in particular,

retired people) play an important role in local economies and help explain the equalization of income across

regions. On these grounds, measure 311 is likely to help reduce poverty and social exclusion among farm-

ers.

However, as noticed above, the indicators used are insufficient to assess the true impact of the measure. This

is problematic as there are some serious potential drawbacks to such subsidies. First, the measure may dis-

proportionately benefit those who are already better off, as they may be more prone to develop new activi-

ties and more able to satisfy the administrative requirements linked to obtaining the subsidies (filing in the

38


FRANCE

appropriate application); second, the measure may be used by those who would have developed the activi-

ty anyway, crowding out those who need the credits more, and having no aggregate impact on diversifica-

tion.

- Measure 312 subsidizes investments made by micro-firms (employing less than 10 persons). The econom-

ic rationale is that small firms in commerce and crafts may accompany the development of the residential

function in rural areas, while providing jobs to local people. Once again, however, there are issues on the

efficient level of subsidies – how large should they be to internalize the positive externalities linked to these

activities? Monitoring the number of aided firms and the total amount of investments does not say much on

the selectivity of the measure (who gets it?) nor on its net effect.

- Measure 313 aims at promoting rural tourism by subsidizing equipment investments and immaterial invest-

ments. The focus is on the quality and the sustainability of the tourism activities developed. Here, the goal

is to leverage the externalities linked to the residential and tourism functions of rural areas. Again, redistrib-

utive and welfare effects cannot been monitored on the sole basis of the indicators used, which only meas-

ure the volume of the aid.

Measures 311 to 313 follow a common rationale: develop employment by subsidizing private and public invest-

ments, so as to better internalize their positive effects on local activities. However, the public funds allocated to

these activities remain quite modest: about 200 million euros, less than 2% of the public funds allocated to the

PDRH.

- Measure 321 targets the development or preservation of basic services to the rural population (health, cul-

ture, employment, transportation).

- Measure 331 provides subsidies (up to 70%) to training and information actions. This measure of capacity

building is likely to be complementary, in the economic sense, with measures 311 to 313. They are howev-

er thought separately.

Overall, measures in axis 3 follow the general policy outlined by the European Council at Lisbonne (2001) in the

sense that they are predominantly active policies to promote employment. One key issue with such measures is to

check that they also reach the most vulnerable members of rural societies. To that end, indicators on the individ-

ual characteristics of the persons effectively reached would certainly be helpful. Sources such as the Observatoiredu développement rural that aims at combining individual information on the recipients’ situation (income) and the

subsidies received will be particularly useful.

Moreover, a general concern with subsidies to economic activities is that they may go to activities that would have

taken place anyway and increase public spending without creating additional activities. Here is an area in which

standard microeconometric experimental and non experimental methods would be necessary to assess the meas-

ures’ net effects.

Two cases of ex post evaluations of rural development measuresAlthough ex post evaluations of rural development policies are rare, some examples lead to interesting, yet some-

what troubling results. We review the results of two different types of evaluations:

- one that makes a global assessment of rural development policies that took place in the EU objective 5b pro-

gram;

- the other that focuses on a specific measure: the fiscal exemption granted to the Zones de revitalisationrurale (ZRR).

Assessing the macroeconomic impact of integrated rural development policies: the case of objective 5b programsin FranceWe base our assessment of objective 5b programs in France on a recent econometric study by Lofredi and Schmitt

(2006). The evaluation concerns programs that took place between 1994 and 1999. However, these programs have

many characteristics in common with the rural development policies planned for 2007-2013.

Three priorities were set to objective 5b rural development programs:

- the diversification of farming and forestry;

- the development of small and medium-sized businesses;

- the development of rural tourism.

39


FRANCE

The allocation of funds was decentralized so as to target the most vulnerable areas at a subregional scale. This was

done within each region by selecting small administrative units (cantons, corresponding to NUTS 4).

Two aspects are of direct interest in order to assess the impact on poverty and social exclusion: how were the funds

allocated? And what was the impact of these funds? The econometric analyses presented by the authors shed light

on the two questions.

- Fund allocation:Eligible areas were selected applying the criteria defined in Regulation (EC)2052/88, that concerned the

level of development, the demographic trends and the share of rural activities. However, these criteria did

not mechanically determine the treated areas, as they did not define explicit eligibility rules (threshold val-

ues on well-defined indicators) and as only imperfect measures of the criteria were available in France at

the NUTS 4 level. In practice, the selection of the eligible areas was rather the results of a negotiation

process involving the Regional assemblies (Conseils régionaux), the structures in charge of European affairs

in the region (the Secrétariats généraux pour les affaires régionales, SGAR) under the monitoring of the

authority representing the State (préfets and sous-préfets). This implies that political, administrative as well

as socio-economic considerations interfered in the selection process.

Overall, the selection process did not lead to a concentrated fund allocation. Within the 18 regions concerned

(out of 22 French regions), 1590 cantons were selected, covering 60% of the land and representing a popu-

lation of 10 million inhabitants. Over the program’s five years, 910 euros were spent per inhabitant (on aver-

age; there was limited variation across regions: standard deviation of 193 euros). This tendency to spread

resources may imply coordination problems, lack of visibility, less leverage and less impact on the most dis-

advantaged areas. Moreover, the statistical analysis shows that the selection criteria were not very closely

followed. Cantons that were close in terms of ex ante socio-economic indicators were not treated identical-

ly, some of them becoming eligible, and some of them not. Technically speaking, the overlap in the support

of the propensity scores of the treated and non treated areas is quite large (see the graphics in Lofredi and

Schmidt, 2006, page 29).

- OutcomeEven though it means that the program was not really focused, this support overlap is an opportunity for

impact evaluation. Non selected cantons with observable characteristics similar to cantons that were select-

ed provide a natural counterfactual to assess the impact of the program. The underlying assumption is that,

had there been no objective 5b program, selected cantons would have performed similarly to similar non

selected cantons. This remains a strong and non tested assumption. It could be violated if, for instance,

regional assemblies were able to target cantons which faced particularly bad economic prospects – in which

case the effect of objective 5b programs will be underestimated by the evaluation.

This notwithstanding, the results are mixed. The most robust effect is that objective 5b programs have

increased the average income per inhabitant. However, there is no clear-cut evidence that they created new

jobs over the 1990-99 period, once preceding trends have been controlled for. Last, the programs were

unsuccessful in inverting demographic trends.

Assessing the employment effects of a targeted tax exemption: the case of the Zones de revitalisation ruralesThe weak aggregate impact of objective 5b programs in France begs the question of whether more targeted micro

measures may have stronger detectable impacts. One such measure may be the tax exemption granted to selected

rural areas. In the short assessment that follows, we build upon the analysis by Lofredi (2006) as well as our own

statistical analyses.

Like enterprise zones in the US, the ZRR program aims at providing incentives for entrepreneurs to set up or

expand their activities in economically less-developed zones. The incentives include regulatory simplifications and

fiscal alleviation on such taxes as payroll taxes.

A key difference with the objective 5b program is that an explicit quantitative criterion was defined for the selec-

tion of areas (again, the cantons) into the measure. Besides secondary criteria, cantons with a low population den-

sity (<31 inhabitants per km?) were targeted. Indeed, as shown by Figure 4 there is a clear discontinuity in the

chances of being eligible to the ZRR program at the population density threshold of 30 vs. 31 inhabitants per km?.

This makes it possible to compare cantons close to the threshold of 31 inhabitants per km?. As these cantons are

similar except for a discontinuous jump in the probability of being eligible into the program, the differences in the

evolution of employment can be safely attributed to the tax exemptions. Results based on this approach confirm

those of Lofredi (2006) that are based on more standard methods: the program did not succeed in raising employ-

40


FRANCE

ment and in attracting firms. A possible explanation is that the measure suffered from underpromotion and lack of

readability; moreover, its criteria might have been too restrictive.

4.2.2 Structural policies

We organize the analysis of structural policies around the recent National Strategic Reference Framework for

Regional Policy 2007-2013 (Cadre de référence stratégique national, CRSN 2006). These orientations are enforced

at a regional level, which may result in some diversity. They are however a good starting point for an ex anteassessment of what could be the impact of structural policies on poverty and social exclusion in rural areas.

Our analysis is done separately for overseas territories (Départements d’outre-mer, DOM) and Metropolitan

France. The former are Convergence regions and will receive around 2.8 billion euros from the structural funds;

the latter are all Competition regions and will receive 9.1 billion euros.

Policies affecting the DOMWith per capita income below 75% of the median EU income, the DOM are classified as Convergence regions.

Their insularity and the distance to metropolitan France make them highly specific. They are predominantly rural

with high poverty rates and a large share of the population depending upon social minima (RMI and other forms

of minimum income).

The key orientations for the DOM detailed in the National Strategic Reference Framework for Regional Policy

2007-2013 (CRSN, 2006) are the following:

1. For competitiveness, the stress is put on the specific transportation and communication needs, involving in

particular public investments in New Information and Communication Technologies (NTIC).

2. For territorial cohesion, the stress is put on urban rather than rural development. The objective is to reinte-

grate suburban districts that have spread without appropriate infrastructures; the infrastructures for rural

areas appear a lower priority.

3. The employment policy follows the same line as the general employment policy, except for an effort target-

ed at improving the bad image of some occupations.

4. Fighting illiteracy is given a high priority.

These orientations are consistent with the diagnostic of these territories suffering from a low integration to the rest

of France. However, they strikingly remain transversal or predominantly urban-oriented policies.

Policies for Metropolitan FranceThe main orientations for employment, human capital development and social inclusion are common to rural as

41


FRANCE

0.2

.4.6

.81

0 20 40 60 80 100Population density (inhabitants per km2)

Share of ZRR elected cantons

Figure 4: Selection process into ZRR

well as urban areas. They are strongly focused upon employment policies. One key feature is the effort to better

coordinate local players such as the public employment service (ANPE) and the unemployment insurance system

(Unédic), local authorities, associations and firms through local employment planning (the Plans locaux pour l’in-sertion et l’emploi, PLIE). However, the statistics reveal that these plans are more frequent in large urban munic-

ipalities (5000 of the 36,000 French municipalities have a PLIE, but they cover 24 million inhabitants).

Some important orientations for competitiveness are more specific to rural areas. In particular, a specific effort is

planned to facilitate the appropriation of and the access to New Information and Communication Technologies

(NTIC) by small and medium-sized firms in rural areas. It is important to relate this policy to the post-schooling

training policies that have come under the responsibility of the French regions. Indeed, economic theory (in par-

ticular literature on the endogenous adoption of new technologies, e.g. Acemoglu, 2002) and empirical analyses

(e.g. Behaghel, Caroli and Walkowiak, 2007) underscore the fact that the adoption of NTIC is endogenous and is

in particular driven by the characteristics of the labour force. Lowering the cost of physical investments in NTIC

is not enough to facilitate adoption; the issue for firms is also to adapt the skills of their labour force. Statistical

analyses show that small firms in urban areas tend to do so by relying upon the external labour market, i.e. by hir-

ing workers in more skilled occupation and/or getting rid of workers in less skilled ones. This is however less of a

possibility for small firms in rural areas that face in less dense labour market. This likely delays the adoption of

new technologies. In that sense, theoretical as well as empirical results confirm the hypothesized complementari-

ty between human capital policies and the investment in communication infrastructures to reduce the technologi-

cal gap between rural and urban areas.

Though this has to be implemented locally by the regions, a specific development of the National Strategic

Reference Framework for Regional Policy 2007-2013 (CRSN, 2006) is devoted to the territorial adaptation of the

policies. Concerning rural areas, three specific orientations are of interest:

1. The diversification of economic activities has a specific focus upon research and development activities, as

well as innovative services to businesses (consulting, training). Interestingly, these have been for long spe-

cific targets of the Zones de revitalisation rurale that we have analysed above. The results do not seem good,

raising the question of whether these activities can be developed at the appropriate scale in rural areas. This

priority thus appears questionable.

2. Strengthening the links between rural and urban areas is one priority. This acknowledges the major role

played by these links, in particular through the residential function of rural areas. This move is consistent

with recent dynamics (e.g. Perrier-Cornet, 2006).

3. The selection of territories is explicitly conditioned by their ability to group in order to produce documents

about their local development strategies. This bottom-up approach arguably increases the level of initiative.

However, existing studies also suggest that more dynamic territories that have the appropriate human

resources may disproportionately benefit from the funds, thus widening the gap between territories (Guérin,

2006).

To summarize, the current orientations for structural funds into rural areas confirm the national trends. However,

the move toward a better integration of the specific needs of rural areas is partially confirmed. This may help

reduce one of the main drawbacks of preceding programs in rural areas: as Guérin (2006) puts it, “Objective 5b

followed by objective 2 programs have used training and insertion policies to induce rural development, with rather

substantial means and realizations, but with limited impacts due to the predominance of a logic of standardized

supply with little territorial variations”.

Concluding remarks on policy assessment• The orientations currently taken by the policies that may affect poverty and social exclusion in rural areas

are characterized by two features:

• A still strong sectoral bias toward agriculture;

• The predominance of the objective of developing employment through increased economic initiatives (con-

sistently with the Lisbonne strategy).

The net impact of these measures on employment in rural areas is not well-known – and this is probably the first

finding of our analysis: the lack of true impact evaluations. One of the rare econometric analyses available sug-

gests that the impact could be quite mixed (Lofredi and Schmitt, 2006). This justifies the stress put by the European

Commission not only in favour of ex ante but also of ex post evaluations of the existing programs. Moreover,

besides taking net job creation as the measure of impact, it would also progressively be necessary to translate this

into an impact on the income distribution.

42


FRANCE

5. Poverty and groups at risk: case study on poor farmers

In year 2003, 12% of the French lived in households with equivalised disposable income below 60% of the nation-

al median. Among French farmers, this relative poverty rate amounted to 22%. Clearly, poverty among farmers is

an important issue to understand poverty and social exclusion in French rural areas.

The case study proceeds in three steps. First, the historical perspective will underscore the role of macroeconom-

ic factors shaping the income distribution in agriculture, hence the risk of poverty among farmers. Second, a more

detailed description of the different forms of poverty will show various micro-economic forces shaping farmers’

poverty. Third, we will consider the impact of public policies and examples of good practices targeted or not at

French poor farmers but that affect them.

The recent history of poverty among French farmersOver the past 50 years (in particular under the influence of the CAP and the French orientation laws passed in 1960

and 1962), the agricultural sector has undergone major structural changes. In particular, the share of farmers in the

labor force has dropped from about 30% in 1955 to less than 5% today. Simultaneously, the number of persons liv-

ing in farming households has been divided by four. The modernization of the sector is clear from a variety of indi-

cators: the share of so-called “professional” farms (growing more than 12 ha in wheat, or the equivalent) has risen

to 60% in 2003; 23% of farms are incorporated. Productivity has improved dramatically.

However, agricultural prices have witnessed a declining trend, with strong fluctuations due in particular to changes

in the CAP. Poverty rates among farmers have probably been closely affected by these average income movements.

Although we lack consistent time series on the distribution of equivalised disposable income of farmers over these

five decades, Jégouzo et al. (1998) conclude that poverty rates among farmers have slightly declined between the

1980s and the 1990s, remaining above the national average. As shown by Table 3, poverty has developed among

farmers between 1997 and 2003, with the poverty rate rising from 19.8% to 22.0%, while the national poverty rate

was declining from 13.4% to 12.0%.

Table 3: Poverty rates in 1997 and 2003 (based on Guillemin and Legris, 2007)

This contrast reflects the fact that the average disposable has decreased by 1.3% per year between 1997 and 2003,

this being mostly due to the falling agricultural revenues (-5.5% per year).

Other indicators have been used, in particular those based on consumption and deprivations. Over a longer time

period, an unusual indicator turns out to reveal interesting phenomena: the rate of men remaining single at age 35

to 54. The information has been available in France since the 1926 population census. In 1926, male farmers do

not distinguishing themselves from other men. In 1954, single men start being strongly overrepresented among

farmers aged 35 to 54. Research has shown how this can be interpreted as a sign of social exclusion of poor farm-

ers who cannot marry women who are reluctant to share their poor living conditions. The phenomenon is persist-

ent; it particularly concerns small farms and less favored areas (in particular mountains: 32% of male farmers aged

35 to 54 are single in Savoie, 29% in Ariège and Lozère).

Forms of poverty among French farmers

a. Dimension and location of the groupAbout 2 million people live in agricultural households in France today. Their weight in rural areas is higher:

depending on the definition of rural areas, they represent about 10% of the population. The spatial distribution of

43


FRANCE

1997 2003 1997 2003

French population 6.9 6.3 13.4 12.0

Farmers 13.0 15.9 19.8 22.0

Source: Insee, Scees, direction générale des impôts

60% threshold50% threshold

agricultural income and poverty depends on many factors, in particular on the presence of natural handicaps. The

distribution of income is clearly less favorable in high mountains (1% of farms), mountains (15%), piedmont (4%)

and other less favored areas (21%).

b. Gender and ageAs discussed above, the group of poor farmers is characterized by an unbalanced sex ratio, with a large share of

single men.

Another relevant gender dimension is the growing share of farming households where the woman works outside

agriculture. The resulting second source of income substantially increases the household’s disposable income: actu-

ally, for such households, the poverty rate is no longer higher than the national rate.

Again, the phenomenon is not evenly spread over the territory: more remote areas (in particular mountainous ones)

lag behind in terms of the share of women working outside the farm. This can be related to the lower density of

local labour markets in these areas: women may have more difficulties finding the jobs that would provide the addi-

tional source of income.

Differentiation across age groups is also strong. Poverty rates used to be high among retired farmers, due to low

pensions. This is no longer the case. Pensioned farmers hold higher disposable income than younger ones in 1997;

the gap increases when the latter experience a sharp drop in agricultural income between 1997 and 2003.

Public policiesFarmers single out as a group that has been targeted by a broad variety of national and European public policies

for years. Maintaining the income of farmers has been a goal of European policies since the Treaty of Rome (1957).

A variety of tools has been used. This session attempts to assess their impact. To what extent is the persistence of

high poverty rates among French farmers a sign of the failure of public policies?

a. The key role of the CAP and rural development policiesAt the macro level, as shown by the first section, the CAP is a major driver of agricultural income. Such micro

analyses as by Jégouzo et al. (1998) show that the initial forms taken by the PAC (the price guarantee mechanism

which implies that subsidies are based on production levels) have had lasting consequences on the distribution of

subsidies, with higher transfers going to high-income farmers.

To some extent, the choice made in France to base the new single payment on the level of subsidies received over

the years 2000-2002 follows the same direction: CAP subsidies are not primarily used as a redistributive tool to

reduce poverty among farmers. Modulation (direct payments exceeding EUR 5,000 a year being reduced by 5%

from 2007 onward) is too limited to induce major redistributive effect. This judgement however needs to be qual-

ified by two considerations.

1. First, even though CAP subsidies are lower for smaller farms, they constitute a significant share of their

income. Jégouzo et al. (1998) argue that these subsidies have been decisive in allowing poor farmers to

remain in the agricultural sector until retirement, and that this helps explain why farms with negative income

before transfers have been able to survive.

This does not necessarily mean that subsidies have had a negative social impact by allowing persistent

poverty and maintaining unsustainable economic activities. Besides equity considerations, subsidies to

small farmers may be efficient to internalize two sources of externality:

The first one is linked to mass unemployment in France: if poor farmers left their agricultural activities, it

is unclear what fraction of them would find a job. The remaining fraction would have to rely upon the unem-

ployment insurance system. The induced social cost has to be weighed against the cost of the CAP subsi-

dies.

The second externality, explicitly taken into account by the CAP reforms, is the environmental externality

linked to maintaining agriculture in less favoured areas (Indemnités compensatrices de handicap naturel,ICHN). This is a justification for the high share of subsidies in the income of farmers working in these areas.

Subsidies to less favoured areas represent a significant share of the CAP second pillar in France. According

to the 2007-2013 Strategic Plan for Agricultural and Rural Development (PDRH, 2007), they represent 63%

of the credits of axis 2 (environnement), i.e. about 30% of total spending. Such amounts have a decisive

impact on the income of farmers in these areas. According to the 2000 agricultural census, they amount to

75% of the agricultural income in mountainous areas and to 110% in other less favoured areas (meaning

44


FRANCE

that, without subsidies, agricultural activities there would produce an average negative income).

2. The first goal of the single payment for farmers introduced by the Luxemburg reform (2003) was to reduce

the trade distortions linked to other types of subsidies. A side effect is to disconnect part of the farmers’

income from market fluctuations. This turns out to have strong implications on poor farming households, as

they are particularly vulnerable to such fluctuations.

There are other measures from the CAP 2nd pillar that impact poverty among farmers. Let us underscore two of

them: measures targeted at younger farmers, and measures in favour of diversification and multiactivity among

farming households.

Younger farmersYounger farmers face a higher risk of poverty than older ones. As analysed in the recent Strategic Plan for

Agricultural and Rural Development (PDRH, 2007), there are many (young) candidates to instalment in agricul-

ture, but they face rising costs (rising price of land and initial investment costs). Subsidized loans and transfers

have been traditionally used to facilitate instalment. They help explain the high ‘survival rates’ among newly

installed French farmers, as compared to the other self-employed: indeed, using a panel extracted from population

Census data, Jégouzo et al. (1998) find that compared to other self-employed occupations (crafts and commerce),

farmers are characterized by extremely low exit rates from agriculture (three to four times less).

In its most recent version, the measure (PDRH, 2007) is not explicitly designed as a redistributive tool: it is acces-

sible to all newly installed farmers aged less than 40 and satisfying some education requirements. However, the

level of the transfer is modulated on the basis of whether the area is more or less favoured, and of whether the farm

development plan is more or less innovative. Moreover, the subsidized loans are redistributive as they contribute

to reduce the effect of the imperfection of credit markets on poorer farmers: due to the lack of collateral, farmers

with less capital find it harder or more costly to borrow on the market.

MultiactivityMultiactivity of farming households is linked to much lower poverty incidence. There is a risk of poverty trap if

multiactivity becomes necessary for having sufficient incomes and, at the same time, a high level of income is nec-

essary to develop multiactivity (to provide the funds needed for initial investments or to pay for the cost of search-

ing a job). As a consequence, measures in favour of multiactivity may be a powerful poverty reduction tool espe-

cially if they are able to reduce the costs and risks faced by poorer farmers attempting to diversify their sources of

income.

The recent French Strategic Plan for Agricultural and Rural Development (PDRH, 2007) activates this lever.

Measure 311 (“diversification”) provides subsidies to finance diversification investments. However, the measure

is modestly funded (EUR 29 millions, compared to the almost 2 billion devoted to less favoured areas).

The level of spending is of course not a sufficient criterion to assess the effectiveness of the measure. The criteri-

on put forward in the French Strategic Plan is the number of recipients. This too is insufficient for at least two rea-

sons: first, the measure may disproportionately benefit those who are already better off, as they may be more prone

to develop new activities and more able to satisfy the administrative requirements linked to the credit (filing the

appropriate application); second, the measure may be used by those who would have developed the activity any-

way, crowding out those who need the credits more, and having no aggregate impact on multiactivity. On such

micro measures as subsidizing multiactivity, better impact evaluations are feasible, e.g. with randomized introduc-

tion of a new type of subsidies, or with variations in the rate of public funding. They could be usefully implement-

ed to get a better sense of the impact of such measures on economic activity and on poverty (see Duflo et al., 2006).

To summarize, CAP related policies have a direct and massive impact on the farmers’ income, hence on poverty

among farmers. However, these policies have not been systematically analysed for their redistributive impact, and

it is hard to assess their impact on poverty globally, especially if one wants to take into account the behavioural

responses of farmers. Some effects are, however, more clearly identified.

In particular, it appears clearly that 1st pillar subsidies (and the recently introduced single payments) as well as 2nd

pillar subsidies to less favoured areas bring a minimal but stable income to poorer farmers. They thus act as an

insurance tool, kept separate from other social insurance tools (like the minimum income), as they also have a role

in internalizing the positive impact of small farmers on the environment (especially in remote areas).

45


FRANCE

Besides direct transfers, a variety of specific policies may impact farmers’ income, but it is harder to assess their

impact on poverty. Here, impact evaluations such as those developed in microeconometrices would be particular-

ly useful.

b. Social policies and the issue of activating rightsThe RMI (Revenu Minimum d’Insertion) plays a central role within French social policies. Introduced in 1988, it

provides a minimum income to households up to a threshold that depends on the family structure. It is in principle

combined with an agreement on the action that the recipient will conduct in order to progressively better integrate

(the so-called contrat d’insertion).

The RMI was introduced in a context of mass unemployment, in a predominantly urban society, and was therefore

mostly targeted at urban, jobless poor households. It was extended in 1992 to the agricultural sector, i.e. to a very

different public: working poor in rural areas. As documented by Perrier-Cornet and Blanc (2000), only a small

share of farming households whose income is below the RMI threshold actually receive it: about 40,000 farms gen-

erated agricultural income below half the minimum wage, but 7,000 farmers received the RMI. Similarly, based

on case studies, Pagès (2004) shows that potential recipients and social workers in rural areas face specific obsta-

cles. He finds cases where the potential recipients is not informed of his rights or refuses to apply by fear of stig-

matisation, and cases where social workers are not informed of people living in really bad conditions. This raises

important questions: does the RMI miss its target among farmers, why and what could be done?

More in-depth statistical analyses show that the probability that a farmer receives the RMI unsurprisingly reflects

such ceteris paribus determinants as the size of the farm, the date of installation (with farmers installed recently

but when they were older being more at risk), and the remoteness of the area (which probably indicates a labour

market effect, since farming households living in more remote areas are less likely to find employment opportuni-

ties outside the farm). However, once these poverty determinants as well as the type of crops are controlled for,

there still are substantial regional specificities. Keeping other things equal, farmers in the south of France are more

likely to benefit from the RMI than farmers in the north of France (Perrier-Cornet and Blanc, 2000). This points

toward the role of the institutional context of the RMI implementation:

The rule of eligibility makes a distinction according to the tax regime of the farm. Those taxed on the basis of their

actual income can only receive the RMI by derogation, which depends on decisions by local public authorities

(préfets). According to MSA (2005), this derogatory access is actually partially random. Moreover, in four depart-ments, Perrier-Cornet and Blanc (2000) find that derogatory access to the RMI is the rule rather than the exception

(75% of farmers receiving the RMI).

The lack of support from leaders of the agricultural sector (in particular, farmer unions) may also have impeded

the diffusion of the measure. The persistence of an important share of poor farmers does not fit in well with the

official vision of a modernized agriculture, promoting the model of medium-size farms.

Last, the fear of stigmatisation may play a role. However, the sociological survey conducted in Perrier-Cornet and

Blanc (2000) shows that things have probably changed: at least, farmers who receive the RMI have overcome this

fear.

Case study conclusion: the specific features of farmers’ povertyPoverty among farmers has highly specific features in the French context. Whereas poverty and exclusion is gen-

erally related to unemployment or underemployment, a substantial share of farmers (22%, according to the most

recent numbers) constitute a group of full-time working poor.

This specificity can be explained by a conjunction of factors:

- the structural decline in the price of agricultural goods;

- the specific adjustment process of the agricultural sector:

- adjustments take place through prices (revenues) rather than through quantities (employment); this comes

into contrast with the overall functioning of the French labour market;

- correlatively, in most cases, farming remains a lifetime job. The reason for that may be strong individual

preference for the agricultural lifestyle, but it may also be related to the guaranteed minimum income pro-

vided by agricultural policies.

- The relative independence with regard to national social policies (as exemplified by the low access to the

RMI).

46


FRANCE

This begs the question of the efficiency of public policies. The main findings are the following:

- The first pillar of CAP probably helps reducing poverty occurrence among French farmers by providing a

safe, though limited, income to all.

- The impact of rural development policies (including those of the CAP second pillar) on poverty among

farmers remains uncertain, but is probably limited.

- The systematic implementation of national social policies remains an issue.

47


FRANCE

6. Conclusion

After having been neglected for decades, the territorial dimension of poverty and social exclusion is now widely

recognized in France. However, rural areas come only second in attracting attention, far behind suburban areas (lesbanlieues).

The main findings of the report have already been summarized in the overview (section 1). We would like to fin-

ish with a series of key questions:

Features of poverty and social exclusion in rural France

The OMC at a sub-national level?Applying the open method of coordination (OMC) at a sub-national level would imply the definition of shared

objectives and common indicators, the redaction of regional reports and their synthesis, as well as an exchange on

good practices.

Some of these elements are already in place among regions in France. In particular, the French rural development

plans (2007-2013) is based on regional plans. There remain national measures – those that are national by scope

or that enable interregional redistribution – but many measures are being decided upon and evaluated at the region-

al level.

Some lessons can be learnt from the 2000-2006 experience. Part of rural development measures were integrated

with European Structural Funds measures within regional documents (the “Document unique de programmation”,

DOCUP). The process appeared to be successful in involving stakeholders in the discussion; however, it remained

complex and unequally efficient (see Lacombe, Vindel and Vollet, 2007). This underscores the cost and the high

requirements of successful coordination: the process takes time and requires energy, information, and administra-

tive capacities. Moreover, the transaction costs that it induces have to be taken into account.

On the use of success stories in fighting poverty and social exclusion in rural areasThis report has searched for evidence on success stories in fighting poverty and social exclusion in rural areas.

Such evidence is hard to find. Moreover, it is not clear whether searching for success stories is always the good

way to learn. A perverse effect can be that administrations and associations promote their actions and mitigate the

problems they have encountered, and from which there is also much to learn.

This being said, the case of Mutualité Sociale Agricole, for instance, is useful to understand how a decentralized

structure can develop synergies by sharing “good practices”. Independent reports have stressed the efficiency of a

project where the national structure developed tools and shared information with local entities (see section 6).

On the local use of Laeken indicatorsSuch indicators as the Laeken ones have a clear value in focussing policy makers’ attention and moving the polit-

ical agenda by making it possible to set clear objectives. In that sense, they have a natural use at the national level,

where key decisions are taken. However, the decentralization of social policies begs the question of developing

these indicators at the regional level. There is a strong demand for it among regional policy makers and stakehold-

ers.

Asking whether we should monitor indicators along the rural versus urban divide is yet another question. Indeed,

this divide is not directly linked with different centres of decision. So its primarily use would be to monitor gaps,

inequality within entities. This might be quite useful. At the national level, comparing rural and urban areas can be

important. Note that this does not require to have data at very low aggregation levels: indeed, representative house-

hold surveys can be matched with different zoning schemes. Fine-tuned indicators can thus be computed, as they

pool rich individual data together (for an example, see Behaghel, 2007).

48


FRANCE

Agricultural and rural development perspectives

How could agricultural and rural development policies better target poverty and social exclusion?A feature of French rural development policies and the CAP is that they are not primarily policies to fight pover-

ty and social exclusion. Symmetrically, many policies targeted at reducing poverty rarely have a specific rural

focus. This does not mean that they have no effect on rural poverty. But, in the two cases, it is fair to say that we

know relatively little on the impact of these policies on rural poverty and social exclusion. The call for better pol-

icy evaluation stems from the finding of this report that many indicators used to monitor the policies are not true

impact indicators, and that the few existing impact evaluations indicate rather poor impacts. We believe that a spe-

cific effort could be made on evaluating these policies. As valid evaluations require heavy investments for local

projects but benefit to all, we argue that such evaluation should be mutualised and funded by higher levels of gov-

ernment.

What is the role of agricultural and rural development policies to fight unfavourable socio-demographic trends?The “sectoral bias” of current French policies is at odds with the declining role of agriculture in rural areas.

However, forestry and agribusiness still have a significant weight in rural employment, and poverty among farm-

ers is a real and persistent problem. Finding the right balance implies to take into account the fact that agriculture

used to be a lifetime employment: diversification needs time and support. On the other hand, in any plausible sce-

nario, it seems that the “residential function” of rural areas will dominate: activities in this area need to be support-

ed most.

Is there an appropriate level of diversity in the set of chosen measures?This is an old debate and unilateral answers seem inappropriate. Indeed, policies need to take into account conflict-

ing goals (economies of scale vs. local adaptation, efficiency vs. equity, incentives vs. insurance).

A plausible solution lies in the subsidiarity principle. It would imply that most funds are used by member states

and regions with limited constraints. These constraints involve the respect of procedures (to ensure good gover-

nance) and to be consistent with general common orientations. This, however, may not be enough. One way to do

so is to give incentives for experimenting on new measures at the local level, with additional funds being allocat-

ed conditionally on setting up credible impact evaluations from which other states / regions may learn. The value

of such practice would be to internalize the benefits of knowledge for the whole community.

Social and regional development policies

Are further efforts needed to take specific features of poverty and social exclusion into account in the design andimplementation of the national, regional and local strategies for social inclusion?Policies impacting poverty and exclusion in rural areas are rarely specifically targeted on rural areas. However,

poverty and social exclusion display some distinctive features in rural areas (section 3): they are less correlated to

long-term unemployment than in urban areas, less concentrated upon lone-mothers families, and often connected

to the lack of decent housing and to bad access to services and employment, due to remoteness. Furthermore, a

specific difficulty of rural areas is the difficulty to find the human resources to develop projects – such as those

valued by the LEADER approach.

Therefore, specific efforts may be needed in two key directions: activating the rights of rural poor or excluded peo-

ple that lack information or fear the stigma of relying on assistance; developing capacity building through educa-

tion and dissemination of information. But this should not be done at the cost of stressing the gap between rural

and urban areas: existing policies appropriately try to strengthen the links between rural and urban areas. This

acknowledges the growing role played by these links, in particular through the residential function or rural areas.

How could regional policies become more oriented towards fighting poverty?Current regional policies in France have three specific orientations concerning rural areas in France (see section

4): a focus on research and development; the promotion of links between rural and urban areas; a selection of ter-

ritories on the basis of their ability to produce documents on their local development strategies.

This calls for two cautionary tales. Concerning the first point, it is unclear whether developing innovative servic-

49


FRANCE

es to businesses (consulting, training) is feasible, as the experience of the Zones de revitalisation rurales suggests.

More modest objectives may therefore be more efficient (e.g. diversification toward new activities along the resi-

dential function of rural territories; industrial policies to maintain employment in the still quite significant rural

manufacturing sector).

Concerning the third point, one needs to be cautious and proceed by steps in demanding more and more elaborate

bottom-up initiatives: existing studies suggest that more dynamic territories that have the appropriate human

resources may disproportionately benefit from the funds, thus widening the gap between territories.

How can social cohesion policy respond to unfavourable socio-demographic trends?The case of the Départements d’Outre-Mer (DOM) is here emblematic. A generous transfer policy is required, with

a focus on infrastructures, education and employment public policies. Emigration is an equilibrium response and

the causes are better fought than the consequences; long-term changes require shifting the equilibrium.

It should however be noted that, overall in France, the situation concerning migrations is quite mixed, with reverse

migrations taking place from cities to rural areas.

Subsidiarity and cooperation

How can awareness be raised and how could the Social Partners be motivated?This report argues that awareness must take two dimensions: recognizing some specific features of poverty and

social exclusion in rural areas (without overstating them) and recognizing that we know little (and invest insuffi-

ciently to learn) on the best policies to fight this poverty and social exclusion.

To raise this awareness, indicators or communication campaigns are not enough. A key lever for better cooperation

and awareness is to collectively invest in actions that adequately evaluated. The interest raised by such recent proj-

ects as the Revenu de solidarité active (RSA) in France (not a specifically rural policy) has to do with this exper-

imental approach. The method may in the long run turn be a source of consensus.

What could be the role of the social economy?Bridging the gap between existing measures and targeted people is a specific difficulty in rural areas, and the social

economy has a role to play here. We believe that the practices of Familles rurales in Lorraine are a good example

(see section 6).

50


FRANCE

7. References

Acemoglu D. (2002) Technical change, inequality and the labor market. Journal of Economic Literature, 40, pp. 7-72.

Behaghel L. (2006), La dynamique des inégalités de revenu en France rurale et urbaine (1984-2002), LEA Working Paper,n°0611, November.

Behaghel L., E. Caroli and E. Walkowiak (2007), Innovation and Skill Upgrading: The Role of External vs. Internal

Labour Markets, PSE Working Paper, n°2007-04.

Bernard Brunhes Consultant (2003), Évaluation qualitative du Plan de Lutte contre la Précarité – PLCP, Rapport à la

Mutualité Sociale Agricole, juillet.

Chassard M. and B. Chevalier (2007), Un large éventail de revenus agricoles, in L’agriculture, nouveaux défis, INSEE,

coll. Références, 2007.

DATAR (2003), Structuration de l’espace rural: une approche par les bassins de vie, report by INSEE (joint with IFEN,

INRA and SCEES) to DATAR, July 2003.

Davezies L. (2001) « Revenu et territoires », in Aménagement du territoire, Rapport du Conseil d’Analyse Economique,

pp. 173-192, Paris, La documentation française.

Davezies L. and E. Korsu (2002), Un essai de géographie de la pauvreté, in Travaux de l’Observatoire national de la pau-vreté et de l’exclusion sociale 2001-2002, Paris, La Documentation française, pp. 167-245.

Détang-Dessendre C. and C. Gaigné (2006), Location and unemployment duration, mimeo INRA.Duflo E., R. Glennerster and M. Kremer (2006) Using randomization in development economics research: a toolkit,

NBER Technical Paper n°TP0333.

European Commission (1999), Guidelines for the ex post evaluation of objective 5(b) programmes, Directorate-General

VI.

Fall M. and D. Verger (2005), « Pauvreté relative et conditions de vie en France », Economie et statistique, n°383-384-

385, December 2005.

Guérin M. (2006), Evolution des objectifs et des principes d’intervention des politiques de développement rural sur le

long terme, Ingénieries, numéro spécial FEADER, pp. 27-37.

Guillaume S. (1999), Les exploitations agricoles à faibles revenus persistants, Economie et statistique, n°329-330, pp.

147-164.

Guillemin O. and B. Legris (2007), De 1997 à 2003, repli du revenu disponible et du niveau de vie des agriculteurs mal-

gré la pluriactivité, in L’agriculture, nouveaux défis, INSEE, coll. Références, 2007.

Guilluy C. and C. Noyé (2004), Atlas des nouvelles fractures sociales en France. Les classes moyennes oubliées et pré-carisées, Paris, Editions Autrement.

Heckman J., R. LaLonde and J. Smith (1999), The Economics and Econometrics of Active Labor Market Programs, in

Handbook of labor economics, vol. 3A, edited by O. Ashenfelter and D. Card, Amsterdam, North-Holland, 1999, p. 1865-2097.

Hirlet P., J. Kaiser and F. Streicher (2004), L’économie solidaire en Région Lorraine: développement local, creation deservices et gestion de l’emploi en milieu rural. Report to the Délégation Interministérielle à l’Innovation et à l’EconomieSociale, http://www.associations.gouv.fr/IMG/rtf/r26.rtf

IGAS (2005) Rapport de la mission “Quelle intervention sociale pour ceux qui ne demandent rien ?” http://www.ladoc-

umentationfrancaise.fr/rapports-publics/064000195/index.shtml

Inra et Insee (1998) Les campagnes et leurs villes, collection Portrait social, Insee, 1998.

INSEE Lorraine (2005) « Val de Lorraine 2004 : l’équilibre du Pays s’affirme », INSEE Lorraine n°38.

INSEE (2007) L’agriculture, nouveaux défis. Coll. Références. Paris, 2007.

Jégouzo G., J.-L. Brangeon and B. Roze (1998), Richesse et pauvreté en agriculture, Paris, INRA-Economica.

Lofredi P. and B. Schmitt (2006), Evaluer les impacts territoriaux des politiques de développement économique géo-

graphiquement ciblées: l’exemple du programme objectif 5b en France, mimeo.Maclouf P. (1986), La pauvreté dans le monde rural, Paris, Editions L’Harmattan et Association des ruralistes français.

Mansuy M. and D. Pallez (2002), Dimensions territoriales de la pauvreté, in Travaux de l’Observatoire national de lapauvreté et de l’exclusion sociale 2001-2002, Paris, La Documentation française.

Maurin E. (2004), Le ghetto français. Enquête sur le séparatisme social, Paris, La République des idées.

Mondot R., A. Le Jan and J.Y. Ollivier (2006), “Le zonage, outil controversé de l’action publique pour le développement

des territoires ruraux », Ingénieries eau-agriculture-territoire, n° spécial FEADER, pp. 123-130.

MSA (2005), L’exclusion en milieu rural. Constats, réflexions et actions de la MSA. Mimeo.

ONPES (2006), Rapport de l’Observatoire national de la pauvreté et de l’exclusion sociale 2005-2006, Paris, La

Documentation française.

NPAI (2006), Rapport sur les stratégies pour la protection sociale et l’inclusion sociale, 2006-2008, Paris, France.

Pagès A. (2004), La pauvreté en milieu rural, Toulouse, Presses Universitaires du Mirail.

PDRH (2007), Plan de développement rural hexagonal 2007-2013, version 9, Ministère de l’agriculture et de la pêche,

Paris, avril 2007.

PDRN (2000), Plan de développement rural national, updated version november 2006, Ministère de l’agriculture et de la

pêche, Paris, France.

51


FRANCE

Péchoux S. (2006), Concentration spatiale de la pauvreté et nouvelles fractures territoriales de la société française al’aube du 21ème siècle, ERIS Working Paper.

Perriet-Cornet P. (2006), La politique française de développement rural : quelle pertinence par rapport aux mutations et

perspectives des espaces ruraux français ? Ingénieries, numéro spécial FEADER, pp. 227-236.

Perrier-Cornet P. and M. Blanc (2000), Pauvreté et RMI dans l’agriculture. Rapport au ministère de l’agriculture et à la

Délégation interministérielle au RMI.

Perrin D. (dir) (2003), Rapport de l’instance d’évaluation des politiques de développement rural, Conseil national de l’é-

valuation, Commissariat général du plan, Paris, La Documentation française.

Rogers S. C. (1995), Natural Histories : The Rise and Fall of French Rural Studies, French Historical Studies, Vol. 19,

n°2, pp. 381-397.

Verger D. (2005), “Avant-propos”, Economie et statistique, n°383-384-385, december 2005,

http://www.insee.fr/fr/ffc/docs_ffc/es383-384-385a.pdf

52


FRANCE

Annex - Examples of Good Practices

53


FRANCE

54


FRANCE

55


FRANCE

56


FRANCE

57


FRANCE

58


FRANCE

59


FRANCE

60


FRANCE

povertyand social exclusion in ruralareas€¦ · by contrast, rural development policiesare often...

Documents